This article was downloaded by: [Temple University Libraries]On: 15 November 2014, At: 15:12Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Cultural TrendsPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/ccut20

A cultural map of the United Kingdom,2003Modesto Gayo‐Cal , Mike Savage a & Alan Wardea CRESC , The University of Manchester , 178 Waterloo Place,Oxford Road, Manchester, M13 9PL, UK E-mail:Published online: 24 Jan 2011.

To cite this article: Modesto Gayo‐Cal , Mike Savage & Alan Warde (2006) A cultural map of theUnited Kingdom, 2003, Cultural Trends, 15:2-3, 213-237, DOI: 10.1080/09548963.2006.10384444

To link to this article: http://dx.doi.org/10.1080/09548963.2006.10384444

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the“Content”) contained in the publications on our platform. However, Taylor & Francis,our agents, and our licensors make no representations or warranties whatsoever as tothe accuracy, completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Contentshould not be relied upon and should be independently verified with primary sourcesof information. Taylor and Francis shall not be liable for any losses, actions, claims,proceedings, demands, costs, expenses, damages, and other liabilities whatsoever orhowsoever caused arising directly or indirectly in connection with, in relation to or arisingout of the use of the Content.

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

A Cultural Map of the UnitedKingdom, 2003Modesto Gayo-Cal, Mike Savage & Alan Warde

This paper employs Multiple Correspondence Analysis (MCA) to map cultural partici-pation and taste in the UK. It constructs what Bourdieu calls a space of lifestyles from

evidence collected in a national random sample survey of the British population in2003. MCA constructs the space relationally on the basis of similarities and differences

in responses to questions about a large number of cultural items in several sub-fieldsincluding music, reading, TV and recreational activity. These items are mapped alongtwo axes and their clustering indicates affinities between tastes and practices across

sub-fields. The cultural patterns are described. We then superimpose socio-demographicvariables, including class, educational qualifications and age, the distribution of which

indicates tendencies for certain categories of person to have shared tastes. The analysisreveals meaningful, socially differentiated patterns of taste. The space of lifestyles

proves to be structured primarily by the total volume of capital (resources) held by respon-dents and by age. Strong oppositions are revealed. An older, educated middle class shares

‘legitimate’ established cultural preferences. The repertoire of a younger middle classgroup contains more contemporary and ‘popular’ items. Less well-educated, workingclass groups are characterised often primarily by lack of cultural participation, but also,

especially among the young, by an aversion to ‘legitimate’ culture.

Keywords: Bourdieu, Cultural capital, Multiple correspondence analysis, Space of lifestyles

Introduction

Other articles in this special issue have shown how individual cultural fields are diff-erentiated in various ways, and that we need to be careful in applying simple ideas

about how ‘popular’ can be distinguished from ‘high’ culture, and so forth. In thisarticle we consider how far we can detect similarities across the cultural fields, so that

we can assess whether lifestyles are organized in systematic ways. It is a central claimof Bourdieu (1984) that there are ‘homologies’ between cultural taste and participation

Cultural Trends

Vol. 15, No. 2/3, June/September 2006, pp. 213–237

Correspondence to: Mike Savage, CRESC, The University of Manchester, 178 Waterloo Place, Oxford Road,

Manchester M13 9PL, UK. Email: Modesto.gayo-cal@manchester.ac.uk

ISSN 0954-8963 (print)/ISSN 1469-3690 (online) # 2006 Taylor & Francis

DOI: 10.1080/09548960600713122

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

across cultural fields. A taste for certain kinds of classical music might thus be associated

with a taste for fine wines, ‘serious’ fiction and art cinema, for instance. However, inrecent years sociologists have argued that specific leisure ‘enthusiasms’ or ‘lifestyles’

increasingly define people’s central life interests (Bellah et al., 1985; Stebbins, 1992).Here, the argument might run that those who love motor racing, for instance, focustheir interests on it to such an extent that they are relatively disengaged from other

cultural pursuits. One might think that the rising importance of the leisure industriesmight be conducive to this kind of specialization, which would suggest that cultural

distinctions in one area might not overlap into cultural distinctions in another(see the discussion in Roberts, 2004). Still further debates have emerged in the wake of

Petersen and Kern’s argument that we are becoming a nation of cultural omnivores inwhich the middle classes increasingly participate in a wide variety of cultural genres

(Petersen & Kern, 1996; see also Savage, Barlow, Dickens, & Fielding, 1992; Warde, 1997).Particular policy concerns overlap with these debates. Current concerns with social

exclusion argue that particular disadvantaged groups are excluded from the social

body, with the implication that most people do indeed partake in many aspects of cul-tural life. As numerous critics point out (e.g. Levitas, 2004), this discourse tends to see

the problem of exclusion as lying on the social margins, and does not recognize thepossibility of entrenched cultural divisions within the social body itself. It is therefore

essential to explore how far we can detect a unitary cultural landscape in Britain,which people of all kinds feel able to partake of according to their specific interests,

or how far, by contrast, cultural life is itself characterized by opposition and divisionin the way that Bourdieu highlights.

This article attempts a preliminary exploration of these issues. There are numerousmultivariate methods which could be used to assess how far there are homologiesacross cultural fields.1 This article is distinctive in using correspondence analysis, a

form of principal components analysis, which allows us to plot people’s cultural pre-ferences in Euclidian space. This is the method that Bourdieu himself used in Distinc-

tion (Bourdieu, 1984), but it remains remarkably under-utilized in Anglophone socialscience. If we treat the resulting figures as cultural maps, it is possible to assess the

extent to which certain preferences and forms of participation are located close to,or apart from, others, so that we can inductively explore the contemporary cultural

landscape. As well as being interesting in its own right, this can also provide a valuablemeans of allowing us to assess how far there appear to be elite cultural forms, and howentrenched cultural inequalities appear to be.

Accordingly, Part 1 reports on how we constructed the variables and operationalizedthem using multiple correspondence analysis (MCA). Part 2 presents and discusses

our cultural map of contemporary Britain, by examining the clustering of cultural pre-ferences and forms of participation. Against those who argue that we are seeing the

proliferation of discrete enthusiasms, we show that there is a very powerful dividebetween the engaged and the disengaged, and evidence also for tension between estab-

lished and popular tastes. Part 3 develops this argument by showing how the distri-bution of social, demographic and economic variables is associated with the space

of lifestyles.

214 M. Gayo-Cal et al.

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

Part 1: Using Multiple Correspondence Analysis

There is a small literature in English on MCA (e.g. Clausen, 1998; Greenacre & Blasius,1994). MCA proceeds differently from many standard exploratory multivariate tech-

niques which seek to define a valid, distinct, dependent variable which might then beexplained through different combinations of independent variables (see for instance,

Chan & Goldthorpe, 2005). Rather, it proceeds inductively from a complex contin-gency table, not by providing summary statistics, but by transforming its values

into co-ordinates on a figure, whose dimensions can then be interpreted. It canthus be seen as a descriptive procedure in the way Abbott (2001) endorses, thoughit should also be emphasized that in its advanced forms it can also be used in associ-

ation with inferential statistics (LeRoux & Rouanet, 2004).In deriving its findings from the contingency table, each individual is placed in a row,

whilst the columns are composed of binary (‘yes’ or ‘no’) responses to a range ofquestions covering various aspects of cultural taste and participation. It is a deliberate

strategy to ensure that these questions cover a diffuse field, so that it becomes anempirical matter to assess how far patterns can be detected amongst the many modal-

ities. Should questions of a particular type dominate, then it is likely that thedistribution of points in the resulting figures will be the artefact of this dominance.

The 168 categorized variables which form the modalities column of our contingency

table are listed in the Appendix, Table A2. Most of these variables will be familiar fromother articles of this special issue, and need no specific elaboration here, though readers

might need to refer to the Appendix to understand the labels used in Figures 1–7.2 Thevariables have been transformed from their original state to allow them to be used in

MCA, as follows. Those questions which ask respondents to rate their liking for genres(for instance for music and reading) on a Likert 1–7 scale have been recoded, where

1–2 is seen as ‘like’, 3–5 as ‘indifferent’ and ‘6–7’ as ‘dislike’. Those questions whichask about favourite genres of TV, film, visual art, sport and eating out are already in

categorical form, and are used in the original state except where the proportionsstating that they like or dislike the given genre fall (roughly) below 5 per cent of thesample. In such cases these categories are either left out or recoded into a bigger cat-

egory. Those questions asking about the frequency of participating in specific leisurepursuits have also been recoded into three categories (high, low, never).

The Appendix shows that 168 active modalities cover a wide diversity of culturalsub-fields, including TV (23 modalities), cinema (20 modalities), reading (25 moda-

lities), music (38 modalities), visual art (23 modalities), eating out (16 modalities) andparticipation in leisure and sport (21 modalities). Our questions mix questions on

participation and taste. This complexity is deliberate: the aim is to see whether,using MCA, there are common structuring properties which operate across these mul-tiple sub-fields, rather than to ‘test’ for a particular relationship hypothesized to exist.

The space we have constructed here is unusually large and rich, and allows us to assessin unprecedented detail the location of a wide number of cultural variables.3

From the binary indicator matrix, MCA will identify the number of axes, each ofwhich separates out the responses along one dimension. By interpreting the eigenvalues

Cultural Trends 215

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

Figure 1 Location of Active Modalities

216M

.G

ayo-Cal

etal.

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

reported for each different axis (eigen values show the amount of variance explained by

each axis), we can assess how many axes offer an adequate summary of the space of life-styles that we have constructed. In our case, as reported in Table A1, we are satisfied

with two axes, although a further two axes, which we do not have the space to reporthere, are also of some importance. Focusing on two axes alone is also economical ofspace since we can portray these as the x and y axes of a conventional figure.4 These

axes operate to separate out responses relationally, vis-a-vis each other, in a way thatcan permit us to assess whether some stand in opposition to others. Thus, if everyone

who liked westerns also liked soap operas, they would be located in the same position.We initially used the French ADDAD and EYELID programme for the analysis, though

we report here comparable findings produced by the Windows-driven package, SPAD.We can subsequently inspect the ordering of this space of lifestyles to determine how

those who have responded in specific ways to any questions are located within it. Bylooking at each question in turn it becomes possible to describe how different kindsof responses are spatially separated, and how they go to make up key components of

the space of lifestyles. The process is inductive: it is up to the viewer to see if he orshe can discern the reasons why responses are differentially located. To repeat: MCA

does not directly test a given hypothesized relationship.Finally, we can explore what appear to be the structuring features of the space

of lifestyles. An important aid here is the use of socio-demographic variables as‘supplementary’ variables. These variables do not themselves help to construct the

space of lifestyles, but can be superimposed on them, so that we can see if theyappear to be associated with such lifestyles. This strategy has the further advantage

that we are able to see how our results compare to those of Bourdieu, whose use ofcorrespondence analysis was central to his analysis in Distinction (Bourdieu, 1984).

Part 2: The Space of Lifestyles: A Cultural Map of Britain in 2003

Figure 1 maps the co-ordinates of each of the modalities which constitute the space of life-styles where these modalities contribute significantly (i.e. above the mean) to the two axes.

The items represented in this and subsequent figures record only those preferences whichcontribute above the mean to the two axes. To aid interpretation of the associations across

cultural fields, modalities concerned with reading are indicated by an arrow pointingdown; music with a plus sign; TV with a circle; film with an arrow pointing up; eating

out with an arrow pointing right; and sport with a square. A glance at Figure 1 revealsthat, on axis 1, most of the likes are on the left-hand side, and many dislikes are on theright-hand side. Yet, since Figure 1 is rather complicated to view because of the

number of modalities reported, we deal with each of the quadrants separately below ata higher level of magnification. Readers should be aware that Figure 1 is important

because it indicates the complete patterning of the space of lifestyles as a whole. Thefurther apart the modalities are, the more exclusive these modalities are with respect to

each other; the closer together they are, the more likely it is for them to be shared.First consider the bottom left quadrant of our space (Figure 2).5 Remembering that

the further along either axis an item is positioned, the more it discriminates from other

Cultural Trends 217

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

Figure 2 Cloud with those Modalities that Make a Contribution to Axes 1 and 2 (Quarter Bottom Left)

218M

.G

ayo-Cal

etal.

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

modalities, we can see, at the farther points of axis 1, a liking for Impressionist painting,

modern literature and French restaurants. There is also indication of interest in going tothe opera (both frequent and occasional attendance is registered), and frequent visits to

museums, art galleries, orchestral concerts and stately houses. A little closer to the inter-sect we see also a liking for biographies, costume drama and stage musicals, and classicalmusic. These tastes go along with owning original works of art and many books. If we

then look at the items with reference to axis 2, we find opera, French restaurants, clas-sical music and orchestral concerts towards the extremes, and Impressionism, biogra-

phies and regular attendance at art galleries nearer to the middle. It seems relativelyeasy to interpret this quadrant as comprising elements associated with ‘traditional’ or

‘established’ high culture, spanning music, visual arts, eating out, leisure and reading.This includes a taste for items learned and transmitted through classical or formal edu-

cation (classical music, modern literature), items which are comparatively expensive(French restaurants and attendance at opera) and collections of paintings and books,which signify both formal education and discretionary income. There are very few dis-

likes recorded (reality TV and the cheapest of eating-out places are the only two). Thisseems to indicate, therefore, not only a positive reaction to a substantial number of

items, but a tolerance towards others, to some degree in accordance with the idea of cul-tural omnivorousness (Peterson & Kern, 1996). There is little sign, therefore, that those

who appreciate these established cultural forms snobbishly reject other, popular,forms—though we should also note that nor do they show any particular interests in

contemporary and popular culture.Look next at Figure 3, which maps the top right quadrant, the antithesis of the

bottom left. It is comprised almost exclusively of dislikes and avoidances. Modalitiesin this cluster include never attending a museum, art gallery or stately home. Dislikesare held for biography, modern literature, classical music, arts programmes, French

and vegetarian restaurants and landscape painting, owning no books and no paintings.It is significant that this quadrant is the location for those watching more than 5 hours

of TV on a weekday, and that the only distinctive positive preferences located in itare for soap operas, watching sport and horror films. The overriding impression of

this particular configuration of activities is one of disengagement from, and even rejec-tion of, many forms of cultural activity. The dislikes are for those items associated with

established high culture, but there are no alternative positive preferences to suggestsome substitute vibrant popular cultural life.

By contrast, we find in the top left quadrant (Figure 4) evidence of a positive evalu-

ation of contemporary cultural forms, some popular, some serious. The pattern of theco-ordinates is probably best described by its distribution on the second axis, where

items with the heaviest weight are, besides playing football and rugby, liking electronicand urban music, liking modern art, science fiction, fantasy and horror books and TV

comedies. Further along the first axis (contributing to Factor 1) can be found going torock concerts and frequent visits to the cinema, and liking rock and heavy metal

music. The modalities here cover music, playing sport, reading and television.There are some, but not many, dislikes: film musicals, landscape painting, watching

golf on TV and country and western music. There is also considerable evidence of

Cultural Trends 219

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

Figure 3 Cloud with those Modalities that Make a Contribution to Axes 1 and 2 (Quarter Top Right)

220M

.G

ayo-Cal

etal.

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

Figure 4 Cloud with those Modalities that Make a Contribution to Axes 1 and 2 (Quarter Top Left)

Cu

ltural

Tren

ds

221

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

tolerance for almost all forms of non-classical music, registered in the map as

expression of indifference (neither liking nor disliking) to modern jazz, world,heavy metal, electronic and urban music. In this syndrome, popular music plays a sig-

nificant role in defining tastes, though eating Indian and Chinese food, watching sportand going to the pub frequently are also significant.

The fourth quadrant (bottom right) is also primarily structured by dislikes and avoi-

dances (see Figure 5). People sharing this bundle of preferences, apart from expressing aliking for country and western music, typically do not appreciate modern popular music.

They dislike rock, world, urban, heavy metal, modern jazz and electronic music, andnever go to rock concerts. They also do not like to watch football, or horror films, and

do not like modern art or reading science fiction and self-help books. They do not goto night clubs or the cinema, and they do not play sport. They rarely, if ever, go to

pubs or restaurants. This suggests some degree of avoidance of the more commercial cul-tural events and aversion to activities and tastes which have more recently becomepopular. Notably, they are not particularly averse to some other forms of recreational

outings—to art galleries or stately homes, for example. Their distinctive positive prefer-ences are for country and western music, watching social sports (the most commonly

reported of these in the survey were darts and snooker), westerns and films of musicals.They also have preferences for traditional foods, liking the inexpensive outlets of fish and

chip restaurants and cafes, and express dislike for Indian and Chinese restaurants. This setof tastes might perhaps best be characterized as traditional, averse to contemporary

popular culture, and of lower class provenance. This is indeed what we see when we con-sider the supplementary variables below. It is older, working-class people with no

educational qualifications who are situated in this quadrant.However, before pursuing this further, let us review some additional and summary

features of these quadrants. As we can see from Figure 1, a strong line of differentiation

exists along axis 1 between likes and dislikes. The right-hand side of the space of life-styles is characterized by dislikes, especially aversion to the taste for established culture

in the top quadrant and for contemporary popular culture in the lower quadrant, withavoidance of participation in commercial culture more obvious in the latter. Compa-

ring the two bottom quadrants, we see evidence of the archaeology of the culturalfoundations of the post-war years, with the differences being between established

high culture and a taste for what is now a rather old-fashioned syndrome of‘popular’ culture: westerns, fish and chips and musicals. With the exception of fishand chips, it is not people with established tastes (in the bottom left quadrant) who

most dislike these items; rather, it is those younger groups in the top left quadrant.But, in general, there are few dislikes among either the established or the contempor-

ary cluster, the difference between the two being more a difference in preferencesfor genres of music and reading. The two quadrants in the top half of the map are dis-

tinguished particularly by the extent of dislike of established culture. Tastes are moreconvergent (scores high on axis 2, but around the mid point of axis 1) around liking to

watch sport on TV, and the liking for electronic and perhaps urban music.Second, there is evidence of homologies of taste across fields. So, not only can we see

that tastes for particular genres of music or writing are distributed unevenly, we can

222 M. Gayo-Cal et al.

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

Figure 5 Cloud with those Modalities that Make a Contribution to Axes 1 and 2 (Quarter Bottom Right)

Cu

ltural

Tren

ds

223

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

also see that there are affinities between tastes and practices of different types. The

clearest example is perhaps the tendency noted above for those who like Impressionistpainting also to like opera and French restaurants. Equally, though, those who most

appreciate modern art also tend to like science fiction books and heavy metalmusic. There are homologies across fields. Given that the two axes explain about75 per cent of the total variance in this space of lifestyles, the likelihood that these clus-

ters are symbolically significant is robust. However, some fields seem less likely to yielddistinctions: types of TV programmes, for instance (though not the amount of time

spent watching TV), are rarely an element of a distinctive cultural clustering, andnor, very often, are genres of film. This is in itself important, though, confirming

the role that these media play in providing some points of cultural convergence forgroups whose tastes might, in other aspects, be sharply divergent.

Third, we can see some differences between the ways in which tastes and partici-pation differentiate among clusters. Our space of lifestyles has been constructedwith reference to both. However, they are not necessarily the same, since expressing

a view about how much one likes a cultural item or form is very different, and forinstance requires fewer material resources, than frequent visits to live events or sites

of cultural interest. Each operates slightly differently, as can be seen when we lookat the way in which the social space is superimposed upon the space of lifestyles.

Part 3: How Far is the Space of Lifestyles Associated with Social Divisions?

Our discussion in Part 2 is deliberately designed simply to examine the relationships

between cultural modalities themselves, without smuggling in assumptions about thekinds of people who may partake of these distinctive cultural forms. However, now

that we have interpreted the axes in Figure 1, we are able to overlay the location ofsome key supplementary variables to indicate whether there is a clear relationshipbetween social position and position on the cultural map. Recall that these socio-

demographic factors were not part of the construction of the space of lifestyles, whichdepended for its structure entirely on the extent to which preferences were distributed.

This gives some understanding to the social structure of the space of lifestyles.As we have seen in other articles in this issue, the Cultural Capital and Social Exclu-

sion (CCSE) survey has extensive information on the economic, social and educationalcharacteristics of its respondents. Table A3 in the Appendix indicates the variables that

we included as supplementary variables here.6 Each is placed on the map at the pointwhich represents the mean co-ordinates of a particular category, for instance male orage 18–24. Figure 6 shows gender, age group, occupational class and highest edu-

cational qualification of respondent, and mother’s and father’s highest educational qua-lification. To aid interpretation, lines are drawn linking three key variables: occupational

class, educational qualifications and age. What is most obvious is that measures of classand education are distributed monotonically along axis 1. The spread for respondents’

educational qualification is marginally greater than that for occupational class. It isinteresting that the modality furthest on the left-hand side of Figure 6 is father’s edu-

cation, which suggests the importance of inherited cultural capital for respondents’

224 M. Gayo-Cal et al.

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

Figure 6 Supplementary Variables with Trajectories

Cu

ltural

Tren

ds

225

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

taste. In general, we can see the inter-related nature of education and occupational class,

and the tendency for higher occupational position to be inter-generationally trans-mitted. Educated fathers and mothers, and equally parents themselves with higher

occupational positions, disproportionately educate their children to higher levels,enabling them to obtain more rewarding jobs.7 In Bourdieu’s terms, axis 1 plots theholding of both economic and (institutionalized) cultural capital, both increasing

together from right to left, and is close to his account where he sees the ‘volume ofcapital’ as being the fundamental measure of axis 1. The second factor, displayed on

axis 2, is less easily characterized, though is predominantly age related. Although thedifference is not very great, men tend to be located above the intersect, and women

below; both are close to the mid point of axis 1. The distribution of age groupsclearly does map onto axis 2: the youngest age group is at the top, the oldest at the

bottom, and the intermediate groups run in sequence. It is, however, also the casethat age groups might be seen as running along a diagonal from top left to bottomright. Were it not for the positioning of the (relatively small) youngest group, it

would be just as plausible to seek to understand age as a structuring feature of the hori-zontal as of the vertical axis. The reason is, of course, sociologically immediately com-

prehensible. Respondents aged 25–44 are, on average, significantly more highlyqualified than older cohorts, and were the beneficiaries of changes in the occupational

structure after de-industrialization propelled the growth of professional and managerialoccupations and the decline of manual work. We are thus in a position now to look at

the ways in which social position and tastes are matched.In order to do this we examine Figure 7, which selects modalities from four key

fields—restaurants, genres of television, music and reading. We make this selectionfor ease of reading the diagram, but these supplementary socio-demographic variablescould be superimposed at exactly the same co-ordinates in Figure 1 (and indeed Figures

2–5). In many ways, cognizance of the structure of the social space makes immediatesense of the patterns manifest in the space of lifestyles. It is the educated middle class

who have the tastes clustering on the left of the map; the further one moves to the left,the more these particular tastes are those of graduates and members of the higher eche-

lons of the professional and managerial service class. Those people are particularlylikely to go to opera, like French restaurants and occasionally go to rock concerts.

They are in privileged positions. So were their parents. It is primarily age that differen-tiates between a liking for established culture and one for contemporary cultural forms.It is the younger, educated middle class who have a taste for rock music, heavy metal

and science fiction. However, both the older and younger cohorts exhibit a highdegree of tolerance—they do not exhibit much by way of dislikes. Indeed, the dislikes

are located along the x axis at a point where respondents tend to be educated to A-leveland have lower managerial and intermediate jobs. Dislike for reality TV, fish and chips,

and country and western music is at its greatest among such groups. Thus we see someindication of the striated patterns of taste, as categories of people symbolically differ-

entiate themselves from other groups.The importance of age, and to some extent gender, can be explored further. The

preferences displayed in the top left quadrant are those of younger people. It is no

226 M. Gayo-Cal et al.

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

Figure 7 (a) Cloud of Modalities for the Fields of TV, Eating Out, Reading and Music

Cu

ltural

Tren

ds

227

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

Figure 7 (b) Supplementary variables, mean points.

228M

.G

ayo-Cal

etal.

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

surprise that younger and better-off people might go to rock concerts and night clubs.

Nor that they like contemporary musical genres. Those chosen are perhaps also partlyconditioned by gender. Men have more of a taste than women for science fiction,

fantasy and horror books and heavy metal. Their well-documented preference forwatching sports programmes is also apparent. Conversely, some of the items thatwe know are more preferred by educated women, modern literature, for instance,

lie below the horizontal axis. Nevertheless, it probably is particularly important thatage is represented on the vertical axis. It suggests that some of the forms typically

thought of as formal and established are less central to the younger middle class.For instance, opera, and maybe especially classical music, appear to be preserves of

the older middle class. Size of book collection is also a function of age, along withliking news and nature programmes. These appear to be items common to older

white-collar groups more generally, and do suggest that there might have been a sig-nificant generational shift. What distinguishes the young middle class from the older ismusic, and to some extent food, TV habits and different forms of ‘going out’.

Consideration of the role of age also helps identify the nature of the culturally dis-engaged group. The correspondence analysis suggests that it is young, poorly educated

males who have fewer cultural attachments. They especially dislike those items inwhich the established and older middle class find their identity (owning books,

French restaurants, classical music, current-affairs programmes). The lower the edu-cational level and the greater the involvement in manual work, the more time is

spent watching TV and the more likely a person is to have no books. The youngerthe person, the more they dislike classical music and arts programmes. This group

is more averse to the established middle-class cultural preferences than to those ofthe older working-class cohorts. Presumably this is a function of educational experi-ence and a learned aversion, born out of discomfort with the practice and curriculum

of the school, to types of culture to which there is no attachment or appreciation.If we compare the items appearing in the bottom right quadrant, a position domi-

nated by elderly, poorly qualified people who had had working-class employment, wesee that their aversions are to the items liked by the young middle class. There is the

occasional exception: the division of opinion on fish and chips and modern jazz occursbetween older people across the class boundary. But most dislikes—for forms of con-

temporary music, science fiction and horror movies—are counter-posed to the tastesof the younger middle class.

Conclusions

Let us draw together some conclusions, but before we do this, we should once again

re-iterate the provisional nature of our findings, and the fact that our correspondenceanalysis reported here may well alter in later iterations. Readers need to be aware that

our findings are dependent on the way that we constructed our space of lifestyles,around selected questions from our CCSE survey: this in no way therefore stands as

a definitive statement of our findings as a whole.

Cultural Trends 229

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

MCA allows the opportunity to array a wide variety of forms of cultural taste and

participation visually, so that we can inspect how they are organized in an economicalway. Readers will see that, for any specific field, the separations displayed in Figures 1–5

are consistent with those reported in other articles in this volume, but the value of theexercise here is that we are able to overlay these fields so that interconnections can beusefully unravelled, in a readily accessible form which does not demand statistical

expertise. Rather than testing for specific kinds of relationships or patterns, we arethereby able to present a wide-ranging cultural map.

In interpreting these figures, we have emphasized a number of points. First, not-withstanding widespread claims about the fragmentation of taste communities,

individualization, and similar claims about contemporary forms of social change,we have revealed clear and marked patterns of differentiation in tastes, many of

which appear rooted in long-term historical patterns, and many of which we wouldprobably have found in earlier decades. Thus to learn that older middle-class peoplelike classical music and younger people like electronic music etc. would seem comple-

tely predictable on the basis of long-standing assumptions about cultural taste. Clus-ters of taste do appear to be associated with social position.

This consideration gives us some general basis for asserting that processes of distinc-tion are still at work in Britain, and that the processes work across a range of activities.

Those with high economic and cultural capital are not only likely to go to the opera,but also to watch less than an hour’s television on a weekday, own books, and like rock

music and modern jazz. These are component parts of lifestyles segmented by posses-sion of economic and cultural capital. And there is a consolidation across fields in

much the way that Bourdieu would anticipate. These are not wholly isolated tastecommunities, to be sure. Many items probably have relatively limited importance indetermining cultural identity and identification. A few indicators (those at the

margins of the space) are particularly useful as signs of social differentiation. Thisdoes suggest that cultural consumption is part of collective social identity and

identification, as Bourdieu argues, and is a function of social institutions and socialconnections.

This having been said, we can see that, whereas amongst the older age groups, at thebottom of Figures 1 and 7, there seems to be an opposition between those who like forms

of established, ‘legitimate’ culture, and those who like what used to be seen as more‘popular’ forms of culture, for the younger age groups, at the top of Figures 1 and 7,this is less clear. For the younger age groups, the major rift is between those who

express likings, especially towards forms of popular music, and those who are disengagedand who do not appear to have any obvious likes. We need to repeat our caution that this

finding depends on the variables used to construct the space of lifestyles, and that it couldbe that those on the top right of Figure 1 have tastes and forms of participation that we

did not ask about. But, as should be clear from the articles in this special issue, we did askabout a wide range of cultural forms. We detect a significant group in the population

which, on the basis of our analysis, seems to be outside the parameters of those culturaltastes that we have measured here. It is overwhelmingly the young, poorly educated

working class who fall into this group. Although we are some way from finding a good

230 M. Gayo-Cal et al.

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

measure of cultural inequalities, it seems likely, on the basis of our findings so far, that this

will overlap other, entrenched inequalities of class and age.Looking at the left-hand side of Figures 1 and 7, we find a second axis between those

who like ‘classic’ established taste and those who like more popular cultural forms: this isa tension within the middle class which primarily sets younger against older respondents.Here our finding on the significance of age in defining taste is important, since in Bour-

dieu’s work the two main axes differentiating the space of lifestyles are, first, volume ofcapital (which we also see in our left–right axis) and, second, type of capital, where cul-

tural and economic capitals are differentiated. We do not find the later differentiation.We would have expected the effects of educational qualification to be located on a differ-

ent axis to those of class if these two forms of capital were clearly separated. This is notthe case. The importance of age is not something that Bourdieu brings out in his work.

This may reflect the fact that few of the questions in his survey appear to be aimed atuncovering age differences (for instance, most of his questions on music appear to bepitched at a middle-aged audience), but certainly in our results we see very powerful

effects of age. This point is interesting in view of the fact that so much recent socialscience has uncovered powerful generational shifts, from Inglehart’s claims about the

rise of post-materialist generations (Inglehart, 1990) to Putnam’s arguments about thefalling off in civic engagement amongst younger age groups (Putnam, 2000).

It may be that this was always the case in the UK, but we doubt it. We think that it islikely that age has emerged as an important differentiating feature, as a result of chan-

ging commercial provision, the development of culture industries more attuned tofashion, youth and profit (especially music), liberalizing education, and stronger

polarization in the economic prospects (especially for men) of younger cohortswith more and less qualifications.

Our final remarks return to the issues of social and cultural inequality. We have

indeed found significant sections of the population, defined largely in terms of levelof education and occupational class, who are relatively little engaged in the cultural

tastes and forms of participation we have asked about. These groups comprise a rela-tively large proportion of our sample; their nucleating points occupy an entire quad-

rant of our space of lifestyles. When those in the bottom right-hand quadrant, whoalso tend to be little involved in any of the forms of reading, art and music that we

inquired about, are added, it seems that a fundamental rift exists between the (multi-ply) engaged and the disengaged. We do not, however, think that the vocabulary of‘social exclusion’ offers a helpful characterization of this situation.

This is because the range of the social groups that are primarily gathered in thesequadrants is far too large numerically, and represents a division that is far too sys-

temic, for them to be construed as marginal populations that are excluded from amajoritarian ‘mainstream’. Equally, we have seen that the ‘mainstream’ is itself signifi-

cantly divided, particularly in terms of age and the ways in which this connects withthe relations between level of education and occupational class position. The corre-

spondence analysis has also demonstrated the complexity of the cultural field, withthe result that we cannot easily construct binary divides between included and

excluded. For if questions focused around the concept of cultural capital are to

Cultural Trends 231

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

enter productively into cultural policy debates concerned with a greater equalization

of cultural life chances, they need to do so in a manner informed by the role it plays, inconnection with other forms of capital, in relation to the whole set of complexly dif-

ferentiated positions within the space of lifestyles that this article has demonstrated.

Notes

[1] For case studies using qualitative data, see Silva (in press).[2] It is unfortunate that we are not able to use colour in Figures 1–7 to aid visual interpretation of the

findings. The tables have been placed on the CRESC website www.cresc.ac.uk so that colour can beused, and so that readers can alter the size of the figures should they wish to inspect them in detail.

[3] Readers should note that the analysis reported here is hence more fully elaborated than thatreported in Savage, Gayo-Cal, Tampubolon, and Warde (2005), and that this account is thereforemore robust, though we will continue to refine the space of lifestyles in future work, to bereported in Bennett et al. (forthcoming).

[4] Additional axes would require additional figures, so complicating the interpretation of the space.[5] Modalities without a label do not contribute significantly to the axis concerned and can hence be

disregarded for our purposes.[6] These amount to only a small proportion of those variables available from the CCSE survey and

our future work will examine the location of these supplementary variables in greater detail.[7] Income, incidentally, is a rather less discriminating variable, but it too maps onto axis 1, with

four income groups running from right to left, from lowest to highest quartile.

References

Abbott, A. (2001). Time matters. Chicago, Il: University of Chicago Press.Bellah, R., et al. (1996). Habits of the heart: Individualism and commitment in American life. Berkeley,

CA: University of California Press.Bennett, T., Savage, M., Silva, A., Warde, A., Gayo-Cal, M., & Wright, D. (forthcoming) Culture,

Capital, Distinction London: Routledge.Bourdieu, P. (1984). Distinction. London: Routledge.Chan, T. & Goldthorpe, J. H. (forthcoming). “Social Stratification and Cultural consumption: music

in England”. European Sociological Review, University of Oxford.Clausen, S. (1998). Applied correspondence analysis: An introduction. Beverly Hills, CA: Sage.Greenacre & Blasius (Eds.). (1994). Correspondence analysis in the social sciences. London: Academic Press.Inglehart, R. (1990). Culture shift in advanced society. Princeton, NJ: Princeton University Press.LeRoux, B., & Rouanet, H. (2004). Geometric data analysis. Dordrecht: Kluwer.Levitas, R. (2004). Let’s hear it for Humpty: Social exclusion, the third way, and cultural capital.

Cultural Trends, 13(2), 41–50.Peterson, R. A., & Kern, R. (1996). Changing highbrow taste: From snob to omnivore. American

Sociological Review, 61, 900–909.Putnam, R. (2000). Bowling alone. New York: Touchstone.Roberts, K. (2004). Leisure inequalities, class divisions and social exclusion in present day Britain,

Cultural Trends 13(2), 57–71.Savage, M., Barlow, J., Dickens, P., & Fielding, A. J. (1992). Property, bureaucracy and culture: Middle

class formation in contemporary Britain. London: Routledge.Savage, M., Gayo-Cal, M., Tampubolon, G., & Warde, A. (2005). Cultural capital in the UK: A pre-

liminary report using correspondence analysis (CRESC Working Paper No. 4). Manchester:University of Manchester. Retrieved from http://www.cresc.ac.uk

Silva, E. B. (in press). Homologies of social space and elective affinities: Researching cultural capital.Sociology, 41(1).

232 M. Gayo-Cal et al.

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

Stebbins, R. A. (1992). Amateurs, professionals and serious leisure. McGill, CA: McGill University Press.

Table A2 Co-ordinates and Contribution of the Active Modalities on the First Two Axes

Axis 1 Axis 2

Variables Modalities Coord. Contr. Coord. Contr.

No. hours TV (an ordinaryweekday)

,1 hour (weekday) 0.45 0.48 20.03 0.002–5 hours (weekday) 0.17 0.28 0.01 0.00.5 hours (weekday) 20.76 2.15 20.02 0.00

No. hours TV (an ordinaryweekend day)

,1 hours(weekend) 0.33 0.19 20.09 0.022–5 hours (weekend) 0.22 0.42 20.08 0.07.5 hours (weekend) 20.50 1.31 0.15 0.16

Type TV programme like best Like News/Current Affairs 0.25 0.17 20.57 1.13Like Comedy/Sitcom 0.11 0.02 0.60 0.82Like Police/detective 20.29 0.09 20.23 0.07Like Nature/History 0.18 0.06 20.47 0.54Like Sport 20.05 0.00 0.66 1.04Like Films 0.14 0.03 0.44 0.35Like Drama 0.35 0.17 20.36 0.24Like Soap opera 20.60 0.96 0.11 0.04

Type TV programme like least Dislike News/Current Affairs 20.49 0.16 0.42 0.15Dislike Quiz/games 0.02 0.00 20.01 0.00Dislike Nature/History 20.49 0.16 0.52 0.23Dislike Sport 20.21 0.11 20.02 0.00Dislike Arts programmes 20.35 0.16 0.74 0.91Dislike Reality TV 0.32 0.43 20.36 0.66Dislike Soap opera 0.19 0.09 20.15 0.08Dislike cookery/gardening/

home decorationprogrammes

20.36 0.11 0.46 0.23

How often do you go to thecinema?

High cinema 0.64 1.09 0.58 1.14Some cinema 0.45 0.98 0.21 0.26Never cinema 20.50 2.00 20.32 1.04

(Table Continued)

Appendix

Table A1 Multiple Correspondence Analysis: Selection of Axis

Axis 1 Axis 2

Eigenvalue 0.1601 0.1272Percentage 5.12 4.02Acumulative % 5.12 9.14Variation (%) 1.10Modified inertia rate (%) 49 28Cumulated modified rates (%) 77

Notes: the modified inertia rate is the squared difference between the eigenvalue of a particular axis and the mean

of all the eigenvalues divided by the sum of all the eigenvalues above the mean. The idea is to stretch out the

Cultural Trends 233

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

Table A2 Continued

Axis 1 Axis 2

Variables Modalities Coord. Contr. Coord. Contr.

Type of films like the best Action/Adventure/Thrillers 0.06 0.01 0.30 0.50Comedy 20.02 0.00 0.35 0.44Costume drama 0.62 0.59 21.16 2.68Documentary 0.11 0.01 20.56 0.47Horror 20.12 0.01 0.95 0.87Musical 20.44 0.20 20.77 0.80Romance 20.28 0.10 0.04 0.00Science Fiction 0.47 0.25 0.54 0.42Westerns 21.04 0.84 20.37 0.13

Type of films like the least Dislike Bollywood 0.24 0.11 0.30 0.22Dislike Costume drama 20.05 0.00 0.66 0.45Dislike Horror 0.02 0.00 20.44 0.77Dislike Musicals 0.05 0.00 0.84 0.97Dislike Romance 20.25 0.04 0.27 0.07Dislike Science Fiction 20.18 0.07 20.41 0.46Dislike War 0.07 0.01 0.09 0.01Dislike Western 0.14 0.03 0.05 0.00

No. books read in last year No book 20.95 2.63 0.24 0.221–6 Books 0.05 0.02 0.20 0.347–24 Books 0.41 0.54 20.21 0.18.24 Books 0.34 0.32 20.50 0.89

How much like whodunnits? Like whodunit 0.14 0.09 0.03 0.00Indifferent whodunit 0.27 0.44 0.05 0.02Dislike whodunit 20.52 1.18 20.16 0.14

How much like Sci-Fi, fantasyand horror?

Like science fiction 0.42 0.40 0.73 1.53Indifferent science fiction 0.31 0.31 0.45 0.86Dislike science fiction 20.21 0.43 20.33 1.33

How much like romances? Like romances 20.24 0.19 20.08 0.03Indifferent romances 0.30 0.45 20.16 0.17Dislike romances 20.11 0.09 0.14 0.17

How much like biographiesand autobiographiess?

Like biographies 0.39 0.93 20.11 0.09Indifferent biographies 0.15 0.13 0.04 0.01Dislike biographies 20.90 2.97 0.09 0.04

How much like modernliterature?

Like modern literature 0.84 1.51 20.25 0.18Indifferent modern literature 0.43 1.12 0.05 0.02Dislike modern literature 20.65 2.86 0.04 0.01

How much like religiousbooks?

Like religious books 0.01 0.00 20.51 0.44Indifferent religious books 0.35 0.42 20.25 0.29Dislike religion 20.13 0.17 0.14 0.27

How much like self-helpbooks?

Like self help 0.33 0.27 20.05 0.01Indifferent self help 0.30 0.45 0.19 0.22Dislike self help 20.30 0.69 20.11 0.13

How often go to Rockconcerts?

High rock concerts 0.80 0.96 0.83 1.31Some rock concerts 0.65 1.36 0.43 0.77Never rock concerts 20.32 1.12 20.25 0.86

How often go to Opera? High opera 0.91 0.59 21.22 1.34Some opera 0.92 1.53 20.47 0.51Never opera 20.19 0.48 0.13 0.27

(Table Continued)

234 M. Gayo-Cal et al.

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

Table A2 Continued

Axis 1 Axis 2

Variables Modalities Coord. Contr. Coord. Contr.

How often go to Orchestral orchoral concerts?

High orchestral concerts 0.76 1.07 21.11 2.85Some orchestral concerts 0.70 1.64 20.26 0.28Never orchestral concerts 20.38 1.47 0.27 0.98

How often go to Musicals? High musical 0.48 0.64 20.41 0.58Some musical 0.46 1.06 20.07 0.03Never musical 20.50 1.89 0.18 0.32

How often go to night clubs? High nightclubs 0.19 0.12 1.11 5.32Some nightclubs 0.38 0.30 0.39 0.39Never nightclubs 20.16 0.26 20.46 2.73

How much like rock? Like rock 0.50 1.03 0.61 1.91Indifferent rock 0.17 0.14 0.15 0.15Dislike rock 20.45 1.20 20.50 1.93

How much like Modern jazz? Like modern jazz 0.37 0.26 20.09 0.02Indifferent modern jazz 0.32 0.63 0.11 0.10Dislike modern jazz 20.37 1.04 20.08 0.07

How much like World music? Like world 0.20 0.07 0.32 0.24Indifferent world 0.39 0.82 0.43 1.28Dislike world 20.29 0.64 20.36 1.29

How much like Classicalmusic, inc Opera?

Like classical 0.37 0.64 20.77 3.53Indifferent classical 0.19 0.20 0.17 0.20Dislike classical 20.60 1.75 0.53 1.75

How much like Country andWestern?

Like country 20.42 0.71 20.44 0.98Indifferent country 0.11 0.07 20.06 0.03Dislike country 0.18 0.17 0.38 0.97

How much like ElectronicDance Music?

Like electronic 0.04 0.00 1.12 2.66Indifferent electronic 0.26 0.25 0.61 1.71Dislike electronic 20.04 0.01 20.38 1.63

How much like Heavy Metal? Like heavy metal 0.46 0.35 0.64 0.85Indifferent heavy metal 0.53 0.83 0.57 1.25Dislike heavy metal 20.21 0.46 20.26 0.89

How much like Urban, inc HipHop and RnB?

Like urban 0.12 0.04 0.97 3.32Indifferent urban 0.33 0.53 0.31 0.63Dislike urban 20.22 0.34 20.55 2.66

How often go to Museums? High museum 0.83 2.56 20.42 0.83Some museum 0.26 0.42 0.07 0.03Never museum 20.85 4.10 0.19 0.26

How often go to Stately homesor historic sites?

High stately homes 0.60 1.85 20.42 1.13Some stately homes 0.09 0.05 0.11 0.08Never stately homes 20.81 3.06 0.31 0.57

How often go to Art galleries? High art galleries 0.96 2.32 20.52 0.87Some art galleries 0.55 1.36 20.06 0.02Never art galleries 20.60 3.01 0.18 0.34

Type of art like the most? Like Performance art 0.19 0.05 0.25 0.09Like Landscape art 20.15 0.18 20.24 0.53Like Renaissance art 0.73 0.34 20.57 0.26Like Still life art 0.09 0.01 0.08 0.01Like Portrait art 20.13 0.02 0.36 0.22Like Modern art 0.27 0.09 0.78 0.99Like Impressionist art 0.97 1.41 20.19 0.07Like no art 21.06 1.51 0.37 0.23

(Table Continued)

Cultural Trends 235

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

Table A2 Continued

Axis 1 Axis 2

Variables Modalities Coord. Contr. Coord. Contr.

Type of art like the least? Dislike Performance 0.40 0.34 0.05 0.01Dislike Landscape art 20.06 0.00 0.89 0.86Dislike Renaissance art 0.05 0.00 0.26 0.06Dislike Still life art 0.28 0.12 0.33 0.22Dislike Portrait art 0.42 0.14 0.29 0.08Dislike Modern art 20.07 0.03 20.34 0.92Dislike Impressionist art 20.13 0.03 20.07 0.01

number of paintings No paintings 20.29 0.78 0.14 0.231–3 paintings 0.28 0.27 20.08 0.03.4 paintings 0.66 1.08 20.44 0.59

How often do you go to thepub ?

High pub 0.08 0.03 0.50 1.45Some pub 0.22 0.33 20.05 0.03Rarely pub 20.52 1.05 20.49 1.22

How often go to somewhere toeat out?

High eat out 0.25 0.61 0.15 0.26Some eat out 20.22 0.21 20.20 0.23Rarely eat out 21.06 1.70 20.37 0.27

Place to eat out like most Like Cafe/Fish&Chip 21.00 1.55 20.21 0.08Like Pizzeria/Italian 0.38 0.41 0.11 0.04Like Pub, Wine Bar, Hotel 20.28 0.37 20.25 0.39Like Indian/Chinese

restaurants0.16 0.13 0.44 1.17

Like French Restaurants 0.96 1.06 20.65 0.63Place to eat out like least Dislike Cafe/Fish&Chips 0.36 0.96 20.29 0.79

Dislike Pizzeria/Italian 20.34 0.09 0.05 0.00Dislike Indian/Chinese

Restaurants20.63 0.95 20.21 0.13

Dislike French Restaurants 20.39 0.12 0.79 0.62Dislike Vegetarian restaurants 20.17 0.10 0.58 1.48

Play favorite to sport Play no sport 20.51 1.78 20.17 0.26Play racquet sports 0.80 0.41 0.43 0.15Play indoor health 0.35 0.40 0.22 0.20Play social sports 20.31 0.04 20.30 0.04Play outdoor healthy 0.34 0.29 20.39 0.47Play football and ruby 0.54 0.24 1.40 2.04Play club and expenses sport 0.59 0.26 0.04 0.00

Type of sport like most towatch

Like Car racing 0.30 0.12 0.31 0.16Like Racquet sports 0.28 0.16 20.57 0.87Like Indoor healthy 0.02 0.00 0.05 0.00Like social sports 20.57 0.85 20.07 0.02Like outdoor healthy 0.23 0.05 20.25 0.08Like football and rugby 0.10 0.06 0.31 0.68Like golf 20.02 0.00 20.65 0.28

Type of sport like least towatch

Dislike Car racing 0.11 0.02 20.34 0.25Dislike Racquet sport 20.36 0.27 0.36 0.34Dislike Indoor health 20.08 0.01 0.47 0.24Dislike social sport 0.28 0.53 20.12 0.12Dislike outdoor health 20.49 0.08 0.19 0.01Dislike football and rugby 20.50 0.44 20.22 0.11Dislike golf 20.02 0.00 0.75 0.69

236 M. Gayo-Cal et al.

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014

Table A3 Frequencies and Coordinates of Supplementary Variables

Axis 1 Axis 2Variables Modalities Frequencies Coord. Coord.

Social class Employers large organization 29 0,85 0,02Higher professional 91 0,88 20,14Lower professional/high technician 237 0,64 20,17Lower managerial 77 0,35 0,14Higher supervisory 72 0,28 20,26Intermediate occupation 192 0,16 20,06Employers small org 105 20,05 20,02Lower supervisory 121 20,28 20,08Lower technician 54 20,38 0,35Semi-routine occupation 311 20,34 0,08Routine occupations 201 20,81 0,10Never worked 41 20,66 0,20DK soc class 3 0,23 1,49

Sex Male 663 0,02 0,24Female 871 20,03 20,19

Education No educational qualifications 444 20,88 20,28GCSE, CSE, O-level, 367 20,07 0,29RSA/OCR Higher Diploma 159 0,19 0,03GCE A-level, Scottish 185 0,39 0,28University/CNAA Bachelor 354 0,85 20,10Other 25 20,31 20,44

Age 18–24 114 20,03 1,3525–34 289 0,26 0,8035–44 307 0,28 0,3045–54 236 0,25 20,1055–64 238 20,04 20,6365–74 186 20,57 20,8875þ 162 20,74 20,90DK 2 20,23 0,17

Ethnic origin White-English 1101 0,04 0,01White-other British 302 20,24 20,05White-Other 42 0,45 20,43Other origin 87 20,07 0,18DK 2 0,17 20,24

Region North 393 0,00 0,06Midlands 269 20,01 20,06Southern England 456 0,05 20,03London 133 0,13 0,08Wales 97 0,05 20,26Scotland 134 20,27 0,13Northern Ireland 52 20,46 20,10

Father’s level of education No education (father) 763 20,23 20,20GCSE (father) 125 0,47 0,57GCE A-level (father) 122 0,53 0,12University (father) 112 1,03 0,08Other education (father) 412 20,20 0,13

Cultural Trends 237

Dow

nloa

ded

by [

Tem

ple

Uni

vers

ity L

ibra

ries

] at

15:

12 1

5 N

ovem

ber

2014